The Voting Rights Act turned 50 on Aug. 6, but the anniversary also doubled as an occasion for voting rights advocates to celebrate a new victory: The day before, the 5th Circuit Court of Appeals ruled that Texas’s 2011 photo ID law was unconstitutional, because it violated the rights of minority voters.
Sen. Cory Booker, D-N.J., went on ABC’s This Week on Aug. 9 to explain why he supported the decision: "Take Texas for example, where Lyndon Johnson's obviously from, they passed these voter ID laws. In the decade before it, 10 years, they only prosecute two people for in-person voter ID, only two people. You're more likely to get struck by lightning in Texas than to find any kind of voter fraud."
We were surprised by the colorful comparison. So we decided to see if we could figure out whether lightning really is more likely to strike in Texas than people trying to cast ballots using fake identities.
In 2014, the National Weather Service gave the odds of being struck by lightning as 1 in 960,000. Their methodology consisted of taking the number of reported deaths and injuries (267) and then adding in another 63 estimated injuries, since not every injury caused by lightning will be properly reported, or even reported at all.
Texas had the second-highest number of lightning-related fatalities from 1959 to 2013, behind only Florida. But that’s mostly because Texas is so big, both in terms of area and population; once you adjust the death rate, Texas actually ranks 33rd.
An expert from the National Weather Service confirmed to us that the probability of being struck by lightning in Texas is slightly lower than the national average, right around 1 in 1.35 million.
So how does this 1 in 1.35 million chance compare to the probability of finding voter fraud?
To find out, we first counted up the total number of ballots cast from 2000 to 2014 using the Texas Secretary of State’s online record. The website includes presidential, congressional, and special elections, but skips over a lot of the more local ones. We came up with just about 71.5 million votes cast, but since the database isn’t comprehensive, we rounded up to 72 million (though that’s probably still an underestimate).
Before delving into how many cases of voter fraud there are, here’s a quick definition. Voter fraud is "the intentional corruption of the electoral process by voters. This covers knowingly and willingly giving false information to establish voter eligibility, and knowingly and willingly voting illegally or participating in a conspiracy to encourage illegal voting by others," according to Lorraine Minnite, a professor at Rutgers and author of the book The Myth of Voter Fraud.
Minnite goes on to distinguish between election fraud — which includes fraud perpetrated by other parties involved in elections, such as advocacy groups or election officials — and voter fraud, which is a subset of election fraud committed specifically by voters.
Booker’s press secretary said he was referring only to in-person voter fraud, or voter impersonation, which is the type targeted by voter ID laws, and given the full context of Booker’s comments, that makes sense. Looking at the kind of fraud that voter ID laws aim to prevent excludes certain categories of fraud, such as those involving absentee ballots.
A spokeswoman for Texas Attorney General Ken Paxton sent us information on electoral prosecutions. Since 2002, there have been a total of 85 election fraud prosecutions resolved. Some of these were dismissed; others were classified as a pre-trial diversion, which generally means that authorities agreed not to charge the defendant.
In total, then, there were 51 guilty or no contest pleas and 9 convictions, for a total of 60 cases of election fraud. Our next step was to identify how many cases were of in-person voter fraud. We reached out to experts who have studied these cases in detail.
Justin Levitt, a professor at Loyola Law School in Los Angeles, said most of the prosecutions would not have been prevented by the voter ID law.
"There were a bunch of prosecutions for unlawfully turning in a ballot that people weren't supposed to have possession. There are a few instances of marking someone else's ballot without their consent — those are all absentee ballots. There are a few of fake registration, or of voting while ineligible — none of which are stopped by a rule requiring ID at the polls," Levitt said. "There are vanishingly few instances of voter fraud — incidents flat-out, not just prosecutions — that could be stopped by applying a rule requiring ID at the polls."
Minnite, the Rutgers professor, pointed us to 2012 court testimony from Major Forrest Mitchell, a criminal investigator with the Texas Attorney General’s office, in which Mitchell admitted that only about five of the more than 300 election fraud referrals that had been investigated since 2002 dealt with in-person voter impersonation. Mitchell’s testimony indicated that only two of these cases could have been prevented by a voter ID law. Three people were prosecuted, though, since one case involved two people. So Booker was off by one when he claimed two people had been prosecuted in the ten years before the ID law was implemented.
Levitt, for his part, found only three credible allegations of fraud in Texas elections since 2000 that could have been prevented by an ID rule. Minnite found four cases of this kind from 2000 until 2014.
Four cases of fraud for 72 million votes makes the chance of voter fraud 1 in 18 million.
Booker said, "You're more likely to get struck by lightning in Texas" than to find in-person voter fraud.
Since 2002, there have been a total of 85 election fraud prosecutions, not all of which have resulted in conviction. Only a small number of those cases -- four or possibly fewer -- included allegations of in-person voter fraud, also called voter impersonation. Those are the kinds of cases the voter ID law in Texas is aimed at preventing, and Booker is right that they have been less frequent than lightning strikes. Overall, we rate his statement True.